Summary. If the state issues too little money, people will try to save in order to acquire their desired stock of money which will result in excess unemployment. In contrast, if too much is issued, that causes excess demand and inflation and the state will need to borrow back some of that money so as to constrain demand. But that involves the poor paying extra tax to fund interest paid to money hoarders, which makes no sense. Ergo the optimum amount of state lilability to issue (i.e. government debt and central bank reserves) is whatever brings full employment while keeping the rate of interest paid on state liabilities at zero.
I’ve no idea what proportion of MMTers will agree with the reasoning here, but the paragraphs below set out my own ideas as to why the optimum rate of interest on state liabilities is zero.
Take an economy where the state (i.e. government and central bank) issue a form of fiat money, as is the norm nowadays. The only significant exception to that norm is the Eurozone, where several countries share a currency and central bank. But actually the conclusions reached below apply to the Eurozone as a whole just as they do to a country which issues its own currency.
What is the optimum amount of currency for the state to issue? Well if it issues an insufficient amount, people will try to save so as to acquire their desired stock of fiat / base money, and that will cause excess unemployment (Keynes’s “paradox of thrift”).
Alternatively, if the state issues too much, that will tend to induce everyone to try to spend away their excess stock of money, which will result in excess demand and excess inflation.
The latter excess demand / inflation can of course be countered by having the state offer interest to anyone who deposits money with the state, and that is known as “national debt”. But that interest has to be funded via tax imposed on the population as a whole. In short, the less well-off pay tax so as to subsidise those who hoard large amounts of money, which clearly makes no sense.
Ergo the best or optimum amount of money is whatever brings about full employment without needing to raise interest on state liabilities above zero.
Of course, given the constant shocks to which economies are subjected to in the real world, it is doubtless impossible in practice to hit the latter ideal all the time or even much of the time. But at least the latter ideal should be the AIM.
Commercial bank money.
Having rather suggested above that the demand for money is met by the state, clearly that is not true in that commercial banks also issue money. However, as is widely appreciated in economics (or at least in MMT circles), commercial bank money nets to nothing: that is, for every dollar of such money issued, there is a corresponding dollar of debt owed by the private non-bank sector to banks. Thus such money is not a net asset as viewed by the private sector non-bank sector. And in addition to there being a demand for commercial bank money, there is presumably a demand for a form of money which is a net asset for the private non-bank sector. The latter form of money is what this article is all about.
As distinct from public debt incurred simply to damp down demand, there is debt incurred so as to fund infrastructure. And clearly since private infrastructure providers incur debt there would seem to be nothing wrong with public infrastructure providers doing the same.
Indeed, there is nothing wrong with that as long as the conditions attached to relevant bonds are the same as those that obtain with public infrastructure: e.g. bond holders stand to lose their money if relevant infrastructure projects fail.
In contrast, if holders of the latter infrastructure bonds are guaranteed against loss gratis the taxpayer, then that’s a subsidy of infrastructure. Of course such a subsidy may be justified, but it’s up to advocates of such a subsidy to provide some very good reasons for the subsidy.
Should infrastructure bonds be QE’d?
Having admitted that government bonds designed to fund infrastructure may be justified, does that mean that having the central bank print new base money and buy back such bonds with a view to imparting stimulus is a good idea?
Well the trouble there is that if it makes sense to fund infrastructure and perhaps other public investments via bonds, it clearly does not make sense to then QE the bonds, because that amounts to funding those investments via new money.
That is, the better course of action would seem to be a more conventional form of stimulus, e.g. simply create more base money and spend it, and/or cut taxes.
Another problem with infrastructure bonds.
The economically untutored think that when government borrows to fund spending, taxpayers who would otherwise have had to fund the spending are excused that sacrifice, and those who lend to government make a sacrifice instead.
The reality is more likely to be that those who lend to government make no sacrifice at all. After all, if you have cash to spare, and find you can lend to government at interest, you’re likely to regard yourself as better off, and thus INCREASE your weekly spending, not cut it. That means the central bank will sniff inflation, assuming the economy is already at capacity, and will thus raise interest rates. That in turn means mortgagors and firms which have borrowed pay more interest, which comes to the same thing as raising tax on middle and lower income / wealth groups: hardly the object of the exercise! In fact that all comes to much the same as a tax increase!
It could be argued against the latter point that only a small proportion of government debt nowadays is bought by individuals as opposed to institutions. However, institutions are owned by individuals, thus the ultimate holders of government debt are still individuals.
Thus if to illustrate, you contribute to a private pension scheme, and the scheme uses spare cash to buy government debt, that makes you better off in much the same way is if you owned that government debt personally. Your consequent increased income / wealth may induce you to cut your contribution to the scheme and spend more on consumer goods. Thus the net effect is arguably much the same as where you personally own government debt.
Plus if we’re going to consider this question on an “all else equal” basis, we need to assume that after tax income inequalities remain constant: i.e. we need to assume government will rectify the above increased inequality by raising tax on the better off and cutting tax on the poor (something that government is in practice quite likely to do).
Plus government will presumably want to make good that increased inequality by raising tax on the rich and cutting tax on the poor. Net effect; much the same as funding the original government spending via tax in the first place!!
Funding infrastructure via government debt is pretty much a waste of time. Infrastructure might as well be funded via tax. All of which leaves us with the conclusion reached several paragraphs above, namely “…the best or optimum amount of money is whatever brings about full employment without needing to raise interest on state liabilities above zero.”
However, if government bonds are in fact used to fund infrastructure, those bonds are clearly very different from bonds issued just to damp inflation. Thus the two types of bond ought to kept separate.
So the final conclusion is that infrastructure bonds apart, interest on government and central bank liabilities ought to be kept at or near zero for as much of the time as possible.